Views > Data visualization > The data-ink ratio (Tables)

“One super table is far better than a hundred little bar charts.”

Edward Tufte

I pity tables. Most of us demote them to the bottom of the data visualization food chain—all too often for the wrong reasons. This is unfortunate because tables could be—in some instances—better than graphs at presenting information.

Tables are best suited for looking up precise values, comparing individual values or presenting values involving multiple units of measure. Graphs, on the other hand, are better for detecting trends, anomalies or relations. In other words, graphs show the forest while tables show the trees.

But lets face it. Tables are badly criticized, insulted and condemned, not because they are inferior data display formats, but because they’re usually poorly constructed. The paragraphs that follow will illustrate techniques and approaches all directed at improving data presentation with tables—in the hope to bring back to tables the reputation they deserve among other modes of data visualization.

Let us look at a badly designed table, as in the heavily encoded example below. The table is based on data from Case Problems In Finance, 10th edition. However, the format is adjusted to reflect the quality of the tables that accompany business presentations or technical reports encountered in the real world. This table shows the amount of short-term credit provided to a multinational company by different banks. Some 286 numbers depict debt exposure of banks to two major types of subsidiaries for the multinational—Finance and manufacturing—and in different countries. It’s hard to see much in this table, beyond that the total amount of debt is around 1.4 bUSD. One needs to put some effort to figure out that CIBC has the highest exposure while Commerzbank is one of the least exposed. The table could clearly convey more than that—had its producer applied some rules or guidelines for table construction.



Have a look at the animated GIF below and see how the above table is being transformed by applying Edward Tufte's concept of the Data-ink ratio.

Here is the final outcome of the transformation. Now we can easily see the total amount of debt. Which bank has the highest exposure, which one the least. The highest concentration of debt by country (i.e. Canada for manufacturing UK for finance).

If the above animation is too fast for you check the slider below and go through the transformation one slide at a time.

The Data-ink ratio Tables 1
The Data-ink ratio Tables 2
The Data-ink ratio Tables 3
The Data-ink ratio Tables 3

Thank you for reading, and I hope you found this useful. If you have any questions, find me on Twitter and ask me anything.